發(fā)布時(shí)間：2018-05-21 03:36

  本文選題：質(zhì)子交換膜燃料電池 + 三維可視化　； 參考：《北京郵電大學(xué)》2017年碩士論文

【摘要】：近年來(lái),隨著人們對(duì)清潔能源越來(lái)越多的關(guān)注,科學(xué)家們對(duì)燃料電池的研究也逐漸深入。燃料電池工作過(guò)程中的水管理是影響燃料電池工作效率的重要因素之一,當(dāng)水量過(guò)多時(shí)會(huì)導(dǎo)致電池內(nèi)部的質(zhì)子交換膜被“淹死”;而水量過(guò)少,則會(huì)導(dǎo)致質(zhì)子交換膜“干死”;這兩種情況都會(huì)使得燃料電池不能正常工作。如何能夠顯式地觀察燃料電池中的水、氣狀態(tài),對(duì)于燃料電池的研發(fā)與改進(jìn)具有十分重要的意義,本課題從圖像處理的角度,以三維可視化圖像的方式來(lái)顯示燃料電池內(nèi)部結(jié)構(gòu),并且在圖像中確定水和氣體的分布狀態(tài)。文中利用CT層析透照技術(shù)對(duì)工作過(guò)程中會(huì)出現(xiàn)的幾種不同水量情況的質(zhì)子交換膜燃料電池進(jìn)行掃描,獲取大量燃料電池層析圖像,用于完成燃料電池內(nèi)部結(jié)構(gòu)的重構(gòu)和呈現(xiàn)。本文首先對(duì)這些層析圖像進(jìn)行了初步的灰度分析與研究,研究發(fā)現(xiàn)燃料電池內(nèi)部不同成分灰度值顯示有較明確的灰度區(qū)間,并且各個(gè)區(qū)間之間有明顯的分割閾值,而這些閾值與灰度區(qū)間可以為后續(xù)的燃料電池成分分割提供很好的分割依據(jù)。由于獲取到的層析圖像為射線圖像,存在噪聲點(diǎn)等影響圖像質(zhì)量的因素,文中重點(diǎn)研究了圖像濾波方法,通過(guò)比較中值濾波、均值濾波、高斯濾波這幾種降噪方法在燃料電池層析圖像中的降噪效果,發(fā)現(xiàn)高斯濾波算法更適用于層析圖像的降噪處理操作。為了更好地進(jìn)行電池內(nèi)部的成分劃分,通過(guò)對(duì)比各種邊緣檢測(cè)算法在燃料電池層析圖像中的檢測(cè)結(jié)果,提出了一種基于Canny邊緣檢測(cè)算子的二維圖像區(qū)域分割方法,利用各成分的灰度閾值在燃料電池層析圖像中對(duì)電池結(jié)構(gòu)以及水和氣體等成分進(jìn)行邊緣檢測(cè)與區(qū)域劃分,通過(guò)對(duì)燃料電池內(nèi)部不同灰度區(qū)域進(jìn)行色彩標(biāo)記,在層析圖像中清晰的判定燃料電池內(nèi)部水和氣體成分。在此基礎(chǔ)上,文章還對(duì)光線投射算法進(jìn)行了優(yōu)化,在新的算法中,背景數(shù)據(jù)被忽略,只需要提取層析圖像分割結(jié)果中燃料電池的目標(biāo)數(shù)據(jù),大大減少了原算法的計(jì)算量,并且通過(guò)對(duì)不同灰度區(qū)間的色彩以及透明度進(jìn)行設(shè)置來(lái)實(shí)現(xiàn)三維圖像的區(qū)域劃分以及透視效果。為了更好地實(shí)現(xiàn)三維效果的顯示,本課題利用VTK (Visualization Toolkit)可視化工具進(jìn)行三維圖像體數(shù)據(jù)的分析以及三維圖像的渲染,同時(shí)采用硬件加速機(jī)制來(lái)提高算法效率。本文將CT圖像三維可視化技術(shù)應(yīng)用于燃料電池工作過(guò)程中水管理的研究中,利用改進(jìn)的圖像分割與三維可視化算法對(duì)獲取到的燃料電池層析圖像進(jìn)行可視化操作,通過(guò)對(duì)圖像不透明度以及色彩的設(shè)置,實(shí)現(xiàn)對(duì)燃料電池內(nèi)部的不同結(jié)構(gòu)進(jìn)行差別標(biāo)記,很好的完成了燃料電池三維圖像的成分分割工作,達(dá)到了在水管理研究中對(duì)燃料電池內(nèi)部成分狀態(tài)分析的要求,并且通過(guò)對(duì)光線投射算法的改進(jìn),在算法效率提高方面效果明顯。
[Abstract]:In recent years, as people pay more and more attention to clean energy, scientists' research on fuel cells has gradually deepened. Water management is one of the important factors that affect the efficiency of the fuel cell. When the water is too much, the proton exchange membrane inside the fuel cell will be drowned, but too little water will lead to the "dry death" of the membrane. In both cases, fuel cells don't work properly. How to observe the state of water and gas in fuel cell is very important for the research and improvement of fuel cell. The internal structure of the fuel cell is displayed by a three-dimensional visual image, and the distribution of water and gas is determined in the image. In this paper, CT tomography was used to scan several proton exchange membrane fuel cells (PEMFC) with different water content in the process of operation, and a large number of chromatographic images were obtained to complete the reconstruction and presentation of the internal structure of the fuel cell. In this paper, we analyze and study the gray scale of these chromatographic images, and find that the gray value of different components of fuel cell has a clear gray range, and there is an obvious segmentation threshold between the different sections. These thresholds and gray levels can provide a good basis for the subsequent fuel cell component segmentation. Due to the fact that the computed image is a radiographic image and there are some factors affecting the image quality, such as noise points, this paper focuses on the image filtering method, and compares the median filtering, the mean filter and so on. The effect of Gao Si filtering on the noise reduction of fuel cell tomography image is discussed. It is found that Gao Si filtering algorithm is more suitable for the noise reduction operation of the tomographic image. In order to divide the components of the battery better, a two-dimensional image region segmentation method based on Canny edge detection operator is proposed by comparing the detection results of various edge detection algorithms in the fuel cell tomography image. The gray threshold of each component is used to detect and divide the edge of the structure of the fuel cell and the components of water and gas in the chromatographic image of the fuel cell, and the different grayscale areas of the fuel cell are labeled with color. The internal water and gas components of the fuel cell are clearly determined in the chromatographic image. On this basis, the ray-casting algorithm is optimized. In the new algorithm, the background data is ignored, only the target data of fuel cell is extracted from the segmentation result of the tomography image, which greatly reduces the computational complexity of the original algorithm. By setting the color and transparency of different gray levels, the region division and perspective effect of 3D images are realized. In order to display the 3D effect better, the visualization tool of VTK Visualization Toolkit is used to analyze the volume data of 3D images and render 3D images. At the same time, the hardware acceleration mechanism is used to improve the efficiency of the algorithm. In this paper, the 3D visualization technology of CT image is applied to the research of water management in the process of fuel cell operation, and the improved image segmentation and 3D visualization algorithm is used to visualize the obtained fuel cell tomography image. By setting the image opacity and color, the different structure of the fuel cell can be marked differently, and the component segmentation of the three-dimensional image of the fuel cell can be completed well. It meets the requirement of fuel cell internal component state analysis in water management research, and through the improvement of ray casting algorithm, the efficiency of the algorithm is improved obviously.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41;TM911.4

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李建秋;方川;徐梁飛;;燃料電池汽車研究現(xiàn)狀及發(fā)展[J];汽車安全與節(jié)能學(xué)報(bào);2014年01期

2 王玉;王宏;康雁;;CT影像中一種基于知識(shí)的脊髓自動(dòng)提取方法[J];儀器儀表學(xué)報(bào);2013年06期

3 張新豐;章桐;;質(zhì)子交換膜燃料電池水含量實(shí)驗(yàn)測(cè)量方法綜述[J];儀器儀表學(xué)報(bào);2012年09期

4 江杰;韓丹;;CT對(duì)泌尿系統(tǒng)結(jié)石成分分析的研究進(jìn)展[J];臨床放射學(xué)雜志;2012年04期

5 王玨;伍立芬;鄒永寧;陶李;徐維;;Zernike矩邊緣檢測(cè)與多項(xiàng)式擬合的CT圖像三維測(cè)量算法[J];儀器儀表學(xué)報(bào);2012年02期

6 王玨;黃蘇紅;蔡玉芳;;改進(jìn)Canny算法的CT圖像環(huán)形偽影校正[J];光學(xué)精密工程;2011年11期

7 王英;曾光宇;;圖像去噪算法研究[J];電腦與信息技術(shù);2011年04期

8 方莉;張萍;;經(jīng)典圖像去噪算法研究綜述[J];工業(yè)控制計(jì)算機(jī);2010年11期

9 嚴(yán)華剛;付璇;錢雅君;黃菊英;李海云;;基于VTK的任意平面CT圖像二維交互分割初步研究[J];中國(guó)醫(yī)療設(shè)備;2010年05期

10 余偉巍;席平;何飛;;利用VTK與MFC的醫(yī)學(xué)模型重建方法研究與實(shí)現(xiàn)[J];工程圖學(xué)學(xué)報(bào);2009年04期

相關(guān)博士學(xué)位論文 前1條

1 李文杰;納米CT三維圖像處理分析方法及其應(yīng)用的研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2011年

相關(guān)碩士學(xué)位論文 前9條

1 方軍;光線投射體繪制算法及鉆孔巖心三維可視化[D];南京理工大學(xué);2014年

2 呂金坤;基于VTK的工業(yè)圖像三維可視化技術(shù)研究[D];中北大學(xué);2013年

3 康頤;基于小波變換的醫(yī)學(xué)CT圖像邊緣檢測(cè)技術(shù)研究[D];成都理工大學(xué);2008年

4 胡戰(zhàn)利;基于VTK的醫(yī)學(xué)圖像三維重建及交互研究[D];哈爾濱工程大學(xué);2008年

5 劉光國(guó);基于GPU的直接體繪制關(guān)鍵技術(shù)研究[D];國(guó)防科學(xué)技術(shù)大學(xué);2007年

6 林松;醫(yī)學(xué)斷層圖像預(yù)處理及三維重建技術(shù)研究[D];燕山大學(xué);2006年

7 許冠軍;數(shù)字圖像去噪算法研究[D];浙江大學(xué);2006年

8 牛翠霞;基于醫(yī)學(xué)CT圖像的體繪制方法研究[D];山東科技大學(xué);2005年

9 趙俊紅;連續(xù)斷層ICT圖像的三維重建與三維顯示的研究[D];重慶大學(xué);2003年

，

本文編號(hào)：1917585